Further reading
Read the source, not the translations.
The book leans on a coherent intellectual tradition. This page is a curated reading list of the load-bearing primaries — the writers, papers, and posts behind the argument. If you finish the book and want to go deeper, start here.
Most mainstream AI commentary is a translation of what these sources already wrote. The translation layer adds noise. This page is the source.
The harness as the operating unit
Read these to land the central frame of the book: that the harness — not the model, not the app — is the surface where serious AI work compounds.
The clearest minimal example of an autonomous research loop. The book lifts its core pattern from this work.
The most credible academic voice on AI-in-the-workplace for an executive audience. Coined the three-tier model / app / harness frame.
The OS analogy (model = CPU, context = RAM, harness = OS) is the clearest teaching frame for non-technical readers.
The 2026 mainstream-engineering legitimization of harness engineering. Useful when you need a respectable source for the term.
The single cleanest expository piece on the harness-as-agentic-moat thesis. If you read one piece on why the harness is the moat, read this one.
The build gap, the failure modes, and the institutional drag
The empirical anchors the book leans on for the gap argument and for why most organizations stall.
The clearest published account of how autonomy compounds inside an agent harness over a long horizon.
The empirical paper behind the gap framing. Hard to dismiss; sourced through real economic data.
The 12x production-rate finding that anchors the chapter on institutional drag. Worth reading the whole report.
The strongest short-form survey of where agentic systems land institutionally. Honest about what fails.
The single best practitioner-press treatment of how agent systems decay without active operating practice.
The builder-leader role and the institutional vocabulary catching up
Sources that document the seat the book argues for, in the language the executive market is reaching for in real time.
Names "Agentic AI Modernization Experience" and promises "fingerprints on the strategy, architecture, and ways of working." Cited in the preface and in Chapters 2, 5, and 9.
The earliest mainstream-press use of the verbatim term "builder-leader" in the executive sense.
Box CEO's public articulation of the same seat from the vendor side. The first public C-suite-tier acknowledgment that the role exists.
The architectural shape of a small operator team producing continuous shippable output. Cited in the team-formation chapter.
The skeptic case, taken seriously
The book engages the strongest version of the AGI-skeptic and LLM-skeptic positions. Read these even if you disagree, and especially if you agree too quickly with the practitioners above.
The most prolific named AGI skeptic. His April 2026 acknowledgment that agentic harness systems are "the single biggest advance in AI since the LLM" is rare, and worth reading carefully.
Raised $1.03B on the "LLMs won't reach AGI" thesis. Career risk gives his skepticism credibility executives can't easily dismiss.
Rigorous on the gap between benchmark performance and real understanding. The strongest academic version of the skeptic case.
The most rigorous AGI-definition voice. Empirical ballast for "current systems are not AGI" without becoming a hater.
For the senior leader writing the next memo
Less technical, closer to how the book's audience actually reads. Useful for sense-checking how operator-frame ideas land with executives who will not read the practitioners directly.
The strongest signal on what is landing with the executive audience. The "onboarding Claudie" series is the cleanest exec-tier articulation of agent onboarding.
UK-based, executive-focused. Writes at the intersection of technology trends and business strategy. The "AutoBeta" piece generalized autoresearch to general-purpose AI work.
Less technical but high reach with media-class executives. Useful for tracking what the enterprise press says about AI capability.
Want a shorter follow list?
The Voices to Follow page is a tighter, names-and-handles version of this list, organized by frame. It's the ten-minute version. This page is the deeper one.
Open Voices to Follow →